ContinuousQueryCache and dropped extend connection

Question

Hi Guys,
It seems that CQC won't automatically recover from a dropped extend connection. This is a problem as we have listeners attached to the CQC which don't get fired if the proxy to which it is connected fails.
CQC also doesn't resync if you access a different remote cache on the same proxy connection.
I've tried adding <heartbeat-interval> to <tcp-initiator> which does nothing.
The only thing that appears to work is to poll cqc.size() on a background thread. At which point it looks from the events received as though the CQC is repopulated form scratch - can you confirm?
Surely there's a better way? If we've added a heartbeat, could we at least have a connection listener so we know if we need to wake up CQCs?
Thanks, Paul

Answer 1

Steps to resolve:
1) Get a reference to your NamedCache using standard coherence API's
2) Create your CQC using the namedCache.
3) Register a member listener for that namedCache so you can detect when the tcp extend connection has gone away, and subsequently returned. When the connection has returned just call cqc.size() to force cqc to reconnect/repolulate.
namedCache.getCacheService().addMemberListener(new MemberListener() {
public void memberJoined(final MemberEvent memberEvent) {
if (memberEvent.isLocal()) {
try{ cqc.size() } catch(Ex....){}
public void memberLeft(MemberEvent memberEvent) {
//could take some action here if needed.
}